| dc.description.abstract | Despite the advances made in the field of lower-limb assistive walking devices, user customization of such devices remains a challenge. This work proposes three novel solutions towards addressing this challenge. Conventional walking controllers for transfemoral prostheses require tedious tuning of 12-20 control parameters per joint. Moreover, these parameters would have to be re-tuned when the terrain’s slope angle changes. The first contribution of this dissertation is a new control framework that develops a set of relationships based on the correlation between the control parameters and the progress in the gait cycle. These relationships, also called joint parameter functions, were determined through data-driven approaches. Implementation of these functions greatly reduced the number of tuning parameters to 3-6 per joint. For the second contribution, this framework was extended to sloped walking by determining a mapping from the slope angle to the necessary joint parameter functions. While these solutions help improve user-customization of prostheses controllers, the mechanical design limitations of assistive devices must also be addressed. The polycentricity of human joints like the knee hinders user customization of assistive walking devices. State-of-the-art knee orthoses mechanisms result in a rotation axis mismatch between the user’s knee and the device. Such mismatch leads to device migration and high interaction forces. The final contribution of this dissertation is a novel self-aligning knee mechanism suitable for a diverse group of users, easing user-customization. | |
